The intelligent driving track has ushered in key specifications. Recently, the Ministry of Industry and Information Technology publicly solicited opinions on the mandatory national standard "Safety Requirements for Intelligent Connected Vehicle Combined Driving Assistance Systems", proposing to use quantitative indicators and clear boundaries to draw a safety red line for intelligent driving assistance systems that have long been in a "fuzzy development" state, ensuring the healthy development of the intelligent connected vehicle industry. Many industry insiders believe that the industry is moving from a "functional competition" to a new stage that emphasizes both security compliance and technological innovation. The personification of technology is more like that of an "old driver" at 9am, which is the busy time of the road. The roads in the Jiuxianqiao area of Beijing are complex, with motor vehicles, non motor vehicles, pedestrians, and other vehicles intertwined at multiple intersections. In such an environment, the reporter participated in the test drive of the Horizon City assisted driving system HSD. After setting the destination, the test vehicle smoothly enters the assisted driving mode and autonomously follows the planned route for about half an hour of smooth and steady driving. When encountering a bus that stops temporarily on the road, the vehicle does not stop in the designated lane to wait, but instead overtakes like an experienced driver. When passing through narrow sections of road with no traffic lights and obstacles on the roadside, vehicles also tend to pass smoothly. Regarding this, Liu Wenyao, the Marketing Director of Horizon Intelligent Driving Products, introduced that HSD has currently achieved a one-stage end-to-end technology. The so-called one-stage approach, in layman's terms, refers to the intelligent driving system being based on a holistic network architecture, rather than multiple system modules being interconnected as before. Liu Wenyao told reporters that the biggest advantage of the one-stage approach end-to-end is to solve the problems of information loss and inconsistency caused by hybrid architecture, improve response speed and achieve low latency. "This is also the reason why we are smoother and more human like when dealing with complex road conditions." "The humanization of intelligent driving" is the focus of current innovation breakthroughs in intelligent driving technology. In addition to end-to-end end-to-end solutions, some car companies are also promoting the use of VLA models to enable intelligent driving to have the ability to understand the world like human drivers, and provide personalized services such as remembering users' speed preferences on specific roads. The transformation of "safety" into quantifiable "hard indicators" technology cannot be achieved without broad industrial prospects. Currently, the global automotive industry is accelerating its transformation towards electrification, networking, and intelligence. From January to July this year, the sales of new passenger cars with combined driving assistance systems in China reached 7.7599 million units, a year-on-year increase of 21.31%, with a penetration rate of 62.58%, an increase of 6.5 percentage points from the previous year. However, for a long time, the concept of assisted driving safety has been relatively vague, and car companies often use vague language such as "L2+" and "L2.99" to promote it. Some companies also abuse the concept of "advanced intelligent driving", blurring the boundary between "driving assistance" and "autonomous driving", and driving accidents are also frequent. Farewell to savage growth, intelligent driving assistance systems will face stronger safety supervision. Recently, the Ministry of Industry and Information Technology publicly solicited opinions on the mandatory national standard "Safety Requirements for Intelligent Connected Vehicle Combined Driving Assistance Systems", which transforms the abstract concept of "safety" into quantifiable and easily detectable "hard indicators". The national standard is scheduled to be implemented on January 1, 2027. The draft for soliciting opinions for the first time breaks down "safety" into 46 sub items such as field tests and 72 hour road tests, and mandates the installation of readable data recording systems to fill the gap in L2 level functional safety technology in China. In terms of applicability, it achieves full coverage of passenger cars, commercial vehicles, as well as basic single lane, multi lane, navigation and other assisted driving systems. In addition, the draft for soliciting opinions also focuses on the issue of "human-machine rights and responsibilities", proposes stricter requirements for driver status monitoring, and clarifies that when the system detects that the hand is off the steering wheel or the line of sight is off the road, it will issue prompts at different time points. Repeated violations will result in a certain period of prohibition from using the combined driving assistance system. After the implementation of the national standard, it will change the competitive logic. It is generally believed in the industry that the implementation of the national standard will change the competitive logic in the field of intelligent assisted driving. While car companies compete for technological innovation and cost reduction, who will comply first and who can continue to comply will become the primary consideration. In the past, some intelligent driving systems may have been able to showcase their performance in one or two scenarios at a press conference. But now, to complete the numerous complex scenario tests required by the national standard, it greatly tests the system's generalization ability. ”In Liu Wenyao's view, the implementation of the national standard will play a positive leading role in the industry. Previously, car companies may have had situations where their technology was not yet mature but they had to 'get on the car first and pay for the ticket later'. However, after the implementation of the national standard, if the technology does not meet the safety bottom line, they will not be able to get on the car first. ”Mr. Zhang, an intelligent driving algorithm engineer, also believes that mandatory national standards have the greatest impact on enterprises that previously relied on "low-cost assembly". They may pay greater costs to meet compliance requirements, thereby exacerbating the process of survival of the fittest. Leading enterprises that have already laid out relevant functional safety systems in advance may amplify their first mover advantage. The trust logic of consumers is also expected to be reshaped. Consumers have always purchased assisted driving functions through car companies' press conferences, sales channels, and other channels to obtain information. In the future, the standardization of national standards will become an important basis for consumers' choices, which will force car companies to accelerate technological compliance. Zhang Hong, a member of the Expert Committee of the China Automobile Dealers Association, believes that L3 level models may become the focus of future market competition, and subsidy policies may be tilted to guide companies to shift their resources towards higher-level assisted driving. (New Society)